4-20 deadband on a HW modutrol motor

Hey all, ive noticed some odd behavior of a honeywell modutrol motor actuating a port valve. When I change the output to the motor by 1%, it doesnt move. Change to by 2%, still doesnt move. Change it by 3%, it moves. Try chaning back by 1 or 2 %, and it wont move. If I change it by 5% and then very quickly change it back 4% (for a 1% total change) it moves to the proper spot. There seems to be some kind of deadband effecting the repsone of the valve. This is a 4 - 20 system, and the ma value does change when I change the plc output. Any ideas?

matt

which actuator motor (model?)

Let's assume the actuator is a recent model Modutrol IV, model M7284

The pdf (scanned, not text, believe it or not) is here:
http://customer.honeywell.com/techlit/pdf/63-0000s/63-2477.pdf

There's an adjustment pot, labeled VR3 appears to be a deadband adjustment.

Here's the manual statement:



Dan

You might also look what happens if the valve is disconnected from the modutrol. If the valve is sticking there could be some undesired results. I have seen that with pneumatic actuators.
I have a new M7284 in my office and will see how it reacts with different VR3 settings.

A minimum 0.16mA setting would correspond to 1% of the 16mA range of 4-20mA. A 0.80mA setting would correspond to 5% of the 16mA range.

The Honeywell text, as usual, is not clear, but the fact that the adjustment setting affects movements in the "opposite direction from the most recent change" seems to be what is happening with regard to your applied changes.

What happens when you provide increasing 1% controller output changes? Does it not move until a 3% output in any given direction is applied?

Dan

Yes, it does not move until I get to 3%. Now, it could be 2.1 or so, as its an interger that controls it; but, it does not move at 2%.
I do not know the model off hand, ill get it tomorrow in the shop. It is a new modutrol tho. Ill check to see if it has a dead band.
As of right now, I added a series of timers and such that every couple seconds backs the singal down 3% for a very brief time and then move the original value back. This has helped, but not as much as a working valve would! Thanks again guys.

It is a 7284C1000. I do not see the pots the manual talks about on it. It said the S1 and S2 and VR4 are only on the q model, but I dont see a VR3 pot anywhere. Any suggestions? Is it only the Q model that does this? Ive placed some calls to our honeywell people, but I cant seem to find anyone at honeywell that actually knows. Thanks

matt

The original link I gave was a scanned pdf, whose document number has no other hits on the web. Being scanned indicates that maybe it was a draft, most probably is no longer current and is obsolete? Who knows?

The current web page for the M728x motors is

http://customer.honeywell.com/Honeywell/UI/Pages/Catalog/ModelCategory.aspx?Catalog=Buildings&Category=M7284_10455&ChannelID=%7B02CD9124-96F3-4A04-8EA7-1777CCCAD5B3%7D

Neither the spec nor product sheet mentions that VR3 adjustment.

So it follows that your motor doesn't have the adjustment.

But it doesn't make sense that a modulating motor cannot provide 1% or 1° resolution changes.

I'll try my Honeywell contact on Monday to see if he can shed any light on your problem, or if he knows who can.

Dan

Ive heard from HW, and the C1000 simply does not have the resolution I need (IIRC, it was 3%). There is another model that is a 160 steps in 90º rotation (c120 I think). I am going to switch the motors out. Thanks again guys. I am surpised that an "industrial" control motor has such poor performace.

matt

Wow. It took 3 calls to find someone willing to research this.

I found a guy at the Honeywell support number 763-954-6464, who had to do 10 minutes of research to find out how many steps of resolution any Modutrol motor would provide (not exactly public information).

The result of his research is that all except 4 models (and all on-off motors) have 55 "repositions", or steps of resolution. Over 90°, this is 1.6° step resolution.

There are 4 'enhanced' Modutrol IV models, all of which are 4-20mA proportional motors, each of which has more than 55 repositions. It isn't clear how many more repositions or how much more step resolution 'enhanced' means.

The 4 enhanced motor models are
M7284C1083
M7284C1091
M7284Q1082
M7284Q1090

While 55 repositions is adequate for typical damper positioning, it is too coarse for serious proportional control of firing rate motors.

Dan

This got me thinking so I did some research and found that none of the manufacturers that I've used (Kromschroder, Landis & Staefa, Belimo, Joventa, Honeywell, Econex, et al) publish data on how many positional steps or the accuracy of their modulating electric actuators.

99% of the time I've used these on easy to control temperature loops where I guess it's not a problem if the actuator only has, say, 55 discrete positions but I have had applications that have been difficult to tune and where the actuator always seems to hunt even when at setpoint.

I would have thought that with the latest technology on servo/stepper motors and encoders (rather than potentiometers) that a 21st century actuator with a positional accuracy of better than 1% should be easy enough to produce at a reasonable price.

Why not an actuator where you could send the requred position and get feedback on the actual position, maybe even alter the speed of travel, all via serial comms?

Maybe such a device exists that I'm unaware of - anybody come across one?

Wow dan, what great info! PLCS.net is more useful than any offical support. Isnt user generated content great! It sure wasnt easy getting it out of honeywell, I am glad you did. We have one of their higher accuracy ones now that will be installed next week. Ill let everyone know how much more accurate it is.
Heres another question then, what acctuators does everyone else use for things such as port valve motors?

I tend to go for Kromschroder GT motors, GT31 usually and GT50 for bigger valves, they do a good range of different supply voltages and travel times and the end of stroke and signalling limit switches are easy to adjust.

Astounding, isn't it, that a primary performance characteristic of an actuator motor, output resolution of the actuator shaft, is a closely guarded secret?

There's a different class of electric actuators for the massive valves used in waste treatment plants, by vendors like Rotork, and I'll bet they're more than willing to define what sort of resolution their actuators provide.
I suspect the nmber of repositions steps doesn't matter a heck of lot when the motor drives a damper in an air handler. And since the major market for this category of actuator is HVAC, well, you take what you get . . .

But I've seen first hand the effect of an acuator that can not resolve sufficient positions to control a gas valve. While investigating a "poor control" complaint, I discovered that Barber Colman motors (this was 20 years ago) used the same die cast housing, same electric motor, same slidewire, but just changed the gear train.

The fast motors were geared lower for speed, the two trade-offs being
- the higher the speed, stop to stop, the lower the torque.
- the faster motors had fewer reposition steps, or further radial movement for each step.
I was never able to clarify why the apparent fewer reposition steps. It would seem that if the motor only runs when the input voltage is applied through the control circuit, the motor could run 0.1 seconds or 0.3 seconds or whatever, until the controller ceases its output, at which time the motor stops.

The complaint was hunting, where the controller couldn't maintain setpoint on a belt oven. The PV oscillation synched with the motor hunting. Backing off the tuning constants to slug the controller response down made no difference, the PV was clearly, to the eye, tracking the motor position and the desired motor position could not be achieved. The ideal cotnrol position was in between motor reposition-steps.

BC's support response was to replace the motor with a slower motor. I did it, it worked. The oven lined out nicely with a slower motor and the motor did not hunt.

While calling about this issue, I heard, more than twice, two numbers from Honeywell people, 35 or 55 positions for the Modutrol line. So bear in mind there's still some question on that value.

Honeywell has a higher torque motor, the Herculine 2000 series, in a different class.

It specs the reposition resolution. There is some interplay between a deadband adjustment in the motor and its resolution, but I can't recall the details.




But it's given its cost, a 2000 can only be justified where the torque and resolution performance is required, like ID or FD control dampers on boilers. This class of actuator can be had with Modbus, as can Rotork.
Dan